Reference Bamber, Hardy and JoughinBamber and others (2000a) reported on the calculation of balance velocities for the whole of the Greenland ice sheet. The balance velocities were calculated using the net mass balance (i.e. accumulation – ablation). Figure 3 (Reference Bamber, Hardy and JoughinBamber and others, 2000a) showed these balance velocities, but erroneously without the ablation term included. Figure 1 (seen here) shows the correct balance velocities for the ice sheet using the net mass balance. Ablation was calculated using a positive degree-day model, as stated in the original paper. The most obvious differences in Figure 1 (compared with the original) occur mainly along the west coast where, for a region between about 67° and 70° N, negative velocities were estimated (black in Fig. 1), indicating a retreating ice front. Elsewhere along the west coast, the flow appears more “channelled” and differentiated into separate flow units.

Fig. 1 Calculated balance velocities over the Greenland ice sheet obtained using the two-dimensional computational scheme of Reference Budd and WarnerBudd and Warner (1996).

The main conclusions are not affected by this error, nor the comparison between balance velocities and GPS-derived velocities. The comparison between balance velocities and interferometric SAR velocities is, however, affected near the margins and particularly for the ice front of Humboldt Gletscher. Here, agreement between the two datasets is substantially better than indicated in figure 4c and d (Reference Bamber, Hardy and JoughinBamber and others, 2000a). In addition, the error affects the comparison between the balance and thermomechanically modelled velocities made elsewhere (Reference Bamber, Hardy, Huybrechts and JoughinBamber and others, 2000b). Again, the main conclusions are unaffected, but better agreement is found in the marginal areas in figure 4 (Reference Bamber, Hardy, Huybrechts and JoughinBamber and others, 2000b). Subsequent analyses (and related publications), using the balance velocities discussed here, have not been affected by the error described above.